A transgene insertion at perinatal lethality (ple) is associated with abnormalities of the cortex

BRAIN RESEARCH ELSEVIER

Brai n Research 727 (1996) 196-204

Research report

A transgene insertion at perinatal lethality (ple) is associated with abnormalities of the cortex David R. Beier ~'*, Holly Dushkin ~, Lisa V. Stone b, Gordon F. Sherman b a Genetics Dit:ision, Brigham and Women's Hospital, 75 l:rancis Street, Boston, MA 02115, USA b N«urology Department, Beth lsrael Hospital, Boston. MA 02115, USA

Accepted 19 March 1996

Abstract In an inbred genetic background, mice homozygous for a transgene insertion at perinatal lethaliO' (pie) were round to be significantly smaller than their heterozygous or wild-type siblings at birth, and rarely survived for more than 48 h. Homozygous progeny of pie mice obtained from a cross with a different strain were viable, but were still not obtained in the expected numbers, demonstrating some deleterious affect of this mutation even in a hybrid genetic background. Homozygous mice demonstrate variable expression of abnormalities in brain development. These usually appear as focal cortical ectopias, but also include other abnormalities, such as polymicrogyria. The genomic sequences corresponding to the region disrupted by the transgene were cloned by isolating a junction fragment between transgenic and wild-type sequences, which was then used to obtain the corresponding region of wild-type genomic DNA. Since some probes from this region do not hybridize with genomic DNA from homozygous pie mice, it appears likely that a deletion event coincided with the transgene insertion. K«vwords: Transgenic insertion mutation: Neuronal migration: Cortical ectopia; Mouse; Perinatal lethali(v: Microgyria: Dysgenesis: Cerebral cortex

1. Introduction The intricacies of mammalian cortical development represent a formidable challenge for biological investigation. A variety of strategies have recently been devised that should serve to increase out understanding of this subject, and, as in other areas of developmental analysis, it is likely that mutations which result in tissue-specific abnormalities will contribute substantially to the elucidation of the genetic and cellular interactions that play a role in brain maturation. There are a number of human heritable disorders that include apparent defects in the migration of cortical neurons as an important component of their disease phenotype. The lissencephalies, including M i l l e r - D i ecker lissencephaly, represent one extreme; there are also disorders with more focal defects, such as the periventricular and cortical heterotopias round in Zellweger's cerebrohaptorenal syndrome [1]. An elevated incidence of local abnormalities of the cortex bas been also been found in

* Corresponding author. Fax: (I) (617) 232-4623; e-mail: [email protected] 0006-8993/96/$15.00 © 1996 Elsevier Science B.V. Ail rights reserved Pli S 0 0 0 6 - 8 9 9 3 ( 9 6 ) 0 0 3 8 0 - 0

patients with developmental dyslexia [11], and intractable epilepsy [10]. In the mouse, there are a small number of well-described disorders of neuronal migration, including the reeler mutation, in which the developing cortical neurons fail to migrate past existing neuronal layers as they normally do in wild-type mice [4,22]; and the New Zealand Black (NZB) and BXSB autoimmune mice, in which neurons migrate through a breach in the external limiting membrane to aberrant positions superficial to cortical layer II [25,26]. We have been studying a transgenic line of mice that has an apparent recessive mutation as a consequence of a transgene insertion event. This putative insertion mutation was originally suggested by out inability to maintain a line of m i t e that was homozygous for the transgene. As the transgene was bred into an inbred genetic background, the effect on viability of the putative mutation was more apparent, and most homozygous mice were found to survive less than 48 h after birth. The locus disrupted in this transgenic line was therefore designated perinatal lethalitv (ple) [2]. Genetic analysis of these mice revealed that the pie transgene is very tightly linked to the mutations microcyti«

D.R. Beier et al. / Brain Resear«h 727 (1996) 196-204

anemia (mk) (1.2 _+ 0.8 cM) and Caracul (Ca) (0.7 _+ 0.7 cM) on mouse chromosome 15 [2]. To insure that the apparent linkage was hOt due to recombination suppression, recombination between pie and the more proximal genetic markers Dominant mega«olon ( Dom) and belted (bt) was also tested, and was found to be entirely consistent with a map position for ple near Ca. Additionally, no evidence of chromosome rearrangement associated with the transgene insertion was evident from physical analysis by in situ chromosome hybridization. The present paper describes the further characterization of the pie mutation using genetic, molecular, and histological analyses. This work confirms the deleterious effect of the homozygous transgene insertion in both inbred and hybrid genetic backgrounds. Moreover, we show that homozygous pie mice show variable expression of neuroanatomical abnormalities that suggest defects in cortical neuronal migration.

2. Methods

2.1. Mite The construction of the transgenic line and its dominant phenotype have been previously described [16,30]. Briefly, fertilized ova of a CD1 × C57BL/6J cross were microinjected with a fusion gene carrying an MMTV promoter placed approximately 1 kb 5' to a mouse c-myc genomic fragment. A founder animal was identified, bred against CD1 animals, and maintained in this outbred background. The transgene was introduced into a C 5 7 B L / 6 J inbred background using mice obtained from The Jackson Laboratory. Analysis in a hybrid genetic background was done using progeny of crosses with C3HeB/FeJLe a / a C a / + mice obtained from The Jackson Laboratory (Bar Harbor, ME).

2.2. Histologv Adult and neonatal mice were anesthetized with ethyl ether and perfused transcardially with 0.9% saline, then 10% formalin for 5 min. Brains were removed and stored in 10% formalin for two weeks. They were then dehydrated in 80%, 95%, 100% ethanol and ethanol/ether. The brains were then embedded in 3% celloidin for 3 - 4 days followed by 12% celloidin for 2 - 3 days. The embedded brains were cul on a sliding microtome at 30 Ixm and every fifth section was stained with cresyl violet for Nissl bodies and mounted on glass slides. In order to process embryos, pregnant mice were anesthetized with ethyl ether, the pups removed, and their heads placed into 10% formalin. The mothers were perfused transcardially and the brains of both the mothers and pups were processed as above. The slides were examined under a light microscope and the presence of cortical microgyria, ectopias, dys-

197

plasias, and other types of brain abnonnalities were noted. Ectopias were judged to be either large, moderately-sized, or small. Large ectopias are characterized by a mushroomlike extrusion of cells into the molecular layer, containing more than 50 neurons; moderately sized ectopias present as collections of neurons in the molecular layer containing between 20 and 50 cells; small ectopias contain less than 20 neurons clustered in layer I. The architectonic and hemispheric location of the anomalies were recorded.

2.3. lmmunohisto«hemistra* The monoclonal antibody Rat-401 (Developmental Studies Hybridoma Bank, NIH, Bethesda, MD) was used to stain radial glial fibers in the embryonic and neonatal braius. Adult tissue was stained for neurofilament fibers using a monoclonal antibody to neurofilament pmtein (68 kDa: ICN Immunobiologicals, Costa Mesa, CA). For both procedures free floating sections were rinsed twice in 0.5 M PBS (pli 7.4) for 5 min. Sections were incubated in 0.6% H,O~ for 20 min to block endogenous peroxidase, and rinsed twice in PBS. Rat-401 was diluted 1:4 in 3 ç normal rabbit serum in PBS with 0.3% Triton X-100, and the neurofilament antibody was diluted 1:250 in 3% normal horse serum in PBS with 0.3% Triton X-100. Semions were incubated overnight on a shaker table at 4°C. The following day, tissue was rinsed twice in PBS. The Rat-401 tissue placed into rabbit anti-mouse immunoglobulins (Dakopatts, Dako Corporation, Carpinteria, CA) diluted 1:20, the neurofilament tissue placed into biotinylated anti-mouse lgG (Vector Labs, Burlingame, CA) diluted 1:60, and both incubated on a shaker table for 2 h at room temperature. Sections were rinsed twice, Rat-401 tissue placed into mouse PAP (Dakopatts, Dako Corporation, Carpinteria, CA) diluted 1:250, ueurofilament tissue placed into ABC (Vector Labs, Burlingame. CA) ruade in PBS and incubated on a shaker table at room temperature for 2 h. Sections were rinsed twice in PBS, and twice in 0.5 M Tris (pli 7.6) for 5 min. Ail tissue was developed in the substrate 0.05% diaminobenzedine (DAB). with ().()1«} H , O , , in 0.5 M Tris (pli 7.6) for about 5 rein or umil proper staining was visualized. Sections were rinsed in Tris, mounted onto chrom-alum subbed slides and dried overnight. Slides were rehydrated and intensified in 0.5c~ cupric sulfate in 0.9% saline |k)r 5 rein, counterstained with methyl green/alcian Nue, dehydrated and coverslipped using Permount.

2.4. Clonin,ç Southern analysis, colony hybridization, restriction analysis and plasmid subcloning experiments were done essentially as described in [24]. DNA sequencing was done using Sequenase (USB, Cleveland, OH) according to the manufacturer's recommendations. The cloning of the junctional fragment containing the transgene and its flanking

198

D.R. Beier et al. / Brain Resear«h 727 (1996) 196 204

A:

+/+

~

~

B: ~

+~

carried a junction fragment between the transgene and a region of presumptive genomic DNA. Sequencing of the subcloned EcoR1/Xhol fragment using a primer derived from MMTV LTR sequence identifies the predicted transgene sequence (with only a few bases deleted from its end) and adjacent sequence that contains mouse repetitive elements. Bacteriophage clones carrying sequences corresponding to the wild-type ple locus were identified in a genomic library cloned in EMBL3. The primers which amplify A63ss, the 239 bp PCR-based clone used in Fig. 1 to demonstrate the apparent deletion of wild-type sequences in pie mice are: Forward: TTCTTTCTCGACAGCTCTCCC Reverse: TTAAGGAGGGGAGTTGTTGGA

Fig. 1. Southern analysis of DNA probes from the ple region. (A) Hybridization of the D15Beil probe (derived from a transgene junction fragment) to Pst l-digested genomic DNA from wild-type ( + / + ), beterozygous ( + / p ) , and bomozygous ( p / p ) transgenic mice. The polymorphic band correlates with the presence of the transgene. (B) Hybridization of an amplified PCR fragment (A63ss) from a bacteriophage clone that is contiguous with sequence of the transgene insertion site to Taq l-digested genomic DNA from wild-type, heterozygous, and homozygous transgenic mice. These sequences are apparently deleted in the transgenic mice (see Section 3).

3. Results

3.1. Genetic analysis of the ple mutation

genomic sequences was complicated by the fact that the transgene is present in 5 - 6 copies and is comprised of MMTV sequences, which are moderately repetitive in the mouse genome, and murine c-myc. Since the M M T V / c myc transgene does not contain an EcoR1 site and is tandemly repeated, unique bands found in Southern analyses of transgenic animals using EcoR1 in combination with other infrequent cutting enzymes should identify junction fragments, with the EcoR1 site being contributed by genomic sequence. Hybridization of a transgene-specific probe to EcoR1/Xhol-digested DNA identifies an 8 kb band not seen in either transgenic or wildtype DNA digested with EcoRl or Xhol alone (data not shown). DNA approximately 6 - 9 kb in size was isolated from an agarose gel and cloned into the Lambda FIX genomic cloning vector (Stratagene). Positive clones were identified by colony hybridization using a transgene specific probe and found to have a 8 kb EcoRl-Xhol insert, which was subcloned. Both the transgene-specific probe and total mouse DNA hybridized with this clone, suggesting it

Two intercrosses were performed to test the affect on viability of the ple transgenic insertion. Heterozygous C57BL/6J p l e / + mice of either the N11 and N13 backcross generations were intercrossed. Determination of the genotype was done using the probe D15Beil (described below) derived from one junction between genomic and transgenic sequences. This permits the discrimination between homozygous transgenic, heterozygous transgenic, and wild-type mice (Fig. 1). Genotype analysis was done on progeny round to be either dead or clearly morbid within two days of birth. The remainder of the litter was genotyped after weaning at three to four weeks. In both crosses there is a marked deficiency of homozygous mice that survive to weaning age; the total of 8 found is about 10% of what would be expected, since an intercross should generate a 2:1 ratio of heterozygous to homozygous offspring (Table l). Homozygous ple mite are over-represented in the cohort of mice found dead or morbid within

Table 1 Genetic analysis of crosses using pie mice Viable progeny

Non-viable progeny

Expt

Cross

Generation

+ / +

pie~ +

pie~pie

+ / +

ple/ +

ple/ple

1 2 Total

ple/+ × ple/+ p l e / + × pie~ +

NI 1

N 13

29 40 69

65 78 143

3 5 8

0 4 4

2 16 18

13 10 23

ple/ +

pie~pie

3

Ca + / + ple × pie~pie

N8

46

21

+ / + and p l e / +

ple/ple

Ca + / + ple X Ca + / + ple

NI7

98

II

Expts I and 2 are intercrosses between C57BL/6J pie~ + mice. Viable progeny were scored at 3 - 4 weeks. Non-viable progeny were either found dead or morbid within 48 h post-birth. Genotype analysis was done using the probe Dl5Beil (Fig. 1). Expts 3 and 4 are crosses using B 6 / C 3 H hybrid mice carrying tbe tightly linked marker Ca in repulsion with ple. Genotypes were inferred based upon phenotype at Ca (see Section 3).

D.R. Beier et al./Brain Resear«h 727 (1996J 196 204

two days of birth (Table 1). The total of homozygous pie mice identified in these populations remains significantly less than the one-half of heterozygotes that would be expected from an intercross; this could be due either to incomplete ascertainment after birth due to maternal consumption of pups, or an in utero effect on viability. Genotype analysis of 35 mice harvested between embryonic day 17 and 19 shows no significant deficiency of homozygous mice, suggesting the effect is post-natal (although the statistical power of this sample to detect a deviation is small). In contrast, heterozygosity ['or pie bas no apparent effect on viability. This has been evident during the ongoing introduction of pie into the C57BL/6J genetic background, which has been done by serial backcrosses. For example, for mice of generations N 14 to N17, backcrosses of ple/+ mice to wild-type females produced a total of 51 pie~+ and 57 + / + mice (X 2 = 0.33, P = hot significant). Since viable adult homozygous mice had been identified for the pic transgenic line when it was derived and maintained in an outbred genetic background [16] we tested whether we could "rescue' homozygous mice by outcrossing pie witb a different strain of inbred mice. This was first done by crossing N8 C 5 7 B L / 6 J ple/+ mice witb C3HeB/FeJLe a / a Ca~+ mice. Ce« is a dominantly aeting coat texture mutation that is very tightly linked to ple (0.7 + 0.7 cM) [2]. It was used as a convenient genetic marker to identify homozygous ple mice, since non-Ca progeny of an intercross between Ca + / + ple mice are bomozygous l'or ple (in the absence of recombination). Numerous viable non-Ca mice were found in the progeny of this cross. Genotype analysis confirmed these mice to be homozygous for the ple transgene. However, tbere is a significant relative decrease in viability even in a hybrid genetic background. This was evident when doubly heterozygous Ca + / + ple mice were backcrossed witb ple/ple homozygous mice. While the ratio of heterozygotes to homozygotes expected from such a cross is 1:1 (based on Mendelian inheritance), less than hall the number of homozygous mice compared to heterozy,,ous~, mice were found at weaning age (X 2 = 13.6, P < 0.001, Table 1). A similar experiment has been repeated witb the same result. NI7 C57BL/6J pie~ + mice were crossed with C3HeB/FeJLe a / a Ca/+ mice and their doubly heterozygous Ca + / + ple progeny were intercrossed. 98 Ca progeny were round. If tbe Ca-appearing

19tJ

mice include both Ca + / + Ca and Ca + / + pie mice in a ratio of 1:2 as expected for an intercross, then 33 homozygous mice should also have been obtained. As shown in Table 1 only 11 were recovered (X 2 = 14.7. P < 0.001), The effect on viability of the homozygous pie transgene is also demonstrated by the f'act tbat we have been able to maintain a line of homozygous mice only witb great difficulty. While both maie and female homozygous mice from tbe hybrid C57BL/6J-C3HeB/FeJLe ( B ó / C 3 H ) genetic background are fertile, their reproductive performance is poor, and females generally have few and small litters. Furthermore, homozygous mice rarely survive more than a year, and appear predisposed to develop a variety of tumors.

3.2. PhenoOT~e of the pie mutation In an inbred genetic background, mice homozygous lk)r the transgene insertion at ple can usually be recognized at birth by their small size. The results in Table 2 show that the average weight in homozygous mice is significantly less than their heterozygous or wild-type sibs. These mice will only rarely have an evident milk spot and frequently appear morbid. Our initial histological analysis revealed that homozygous newborn mice often bave variable accumulation of hepatic microvesicular fat. However. this finding proved inconsistent, and could occasionally be seen in heterozygous sibs. The origin of this finding is unclear, and may represent the incomplete mobilization of hepatic fat stores after birth. Examination of the brain revealed striking abnormalities including ectopic collections of neurons in neocortical layer I, cortical microgyria, focal gliotic areas in the cortex, and partial agenesis of the callosum. These anomalies were almost exclusively present in homozygous mice (Table 3). Tbe most common anomaly was neuronal ectopias in the molecular layer of the cortex. Twenty-seven of 116 (23c~) ple bomozygotes had large ectopic collections of neurons in layer I of the neocortex. The ectopias were characterized by mushroom-shaped protrusions of neurons into the molecular layer and overlying subarachnoid space (Fig. 2). Cortical layers I1-VI underlying the ectopias also were distorted. Usually one or two ectopias were seen per affected brain and most were present in motor areas (Frl, 2, or 3; architectonic locations according to Zilles [32]). The next most affected architectonic area

Fable 2 Weight of newborn mice from pie crosses

+/ +

n

pie~ +

n

pie~pie

il

!.28 _+ 0.14

17

1.29 _+ 0.15

33

1.16 + .09

14

Pn)geny of intercrosses between ( ' 5 7 B L / 6 J p i e / + mice were identified on the day of birth. Mean wcight + slandard deçiation is ~hown in grains. Genotype analysis was donc as described in Fig. l. There is no significant difference between the means ~~f" + / + and pt«' F micc using a t lest. Thc mean weighl of thc pie~pie micc is significantly less ( P < 0.004).

I).R. Beier et al./Brain Research 727 (1996) 196-204

200 Table 3 Incidence of cerebral anomalies in ple mite Homozygous pie

Heterozygous pie

Age

ectop

microgyr

NFZ

agenesis

n

ectop

microgyr

NFZ

agenesis

Ji

adult nb EI7 19 EI3-16 ail

II 8 7 1 27

0 2 0 0 2

1 0 0 0 1

3 0 0 0 3

51 25 19 21 116

2 2 0 0 4

0 0 l) 0 0

0 1 0 0 1

0 0 0 0 l)

27 25 13 16 gl

Abbreviations: ectop, ectopias: microgyr, microgyria: NFZ, neuron-free zone; agenesis, agenesis of the corpus callosum; nb, newbom; E, embr3«mic day.

was Pari. A small number of ectopias also were seen in Cgl, FL, and Par2. When the pie homozygous mice were analyzed by age, it was seen that 22% of the adults (n = 51), 32% of the newborns (n = 25) and 37% of the E17-19 fetuses (n = 19) had cortical ectopias. In addition, one ectopia was seen in a E15 fetus. As in previous studies of NZB mice [27], staining of radial glial fibers in newborns using the Rat-401 antibody revealed a breach in the pial-glial boundary underlying the ectopia (Fig. 3). Neurofilament positive fiber

bundles also were present in the subjacent layers of ectopias in adults (Fig. 3). In addition, 2 cases of microgyria were present in the frontal/motor cortex on the right side (Fig. 4) and one adult had a focal gliotic region in the cortex. Three adult cases of partial agenesis of the corpus collosum also were seen, two of which were in mice with ectopias. Wild-type litter mates had no signs of brain anomalies and only 4 of 81 (5%) heterozygous mice had cortical ectopias. In addition, one newborn heterozygote had a

Fig. 2. (A-D) Digitized composite (in Adobe Photoshop) showing Nissl staining of a large ectopia (arrows) in the left hemisphere of a fetal pie homozygous mouse. Panels are arranged from rostral to caudal with 150 ~m between sections. Processing artifacts were digitally excise& Bar ~ 250 ~tm.

D.R. Beier et al./Brain Research 727 ( 1996i 196 204

focal gliotic region in the cortex. No instances of microgyria or agenesis were seen. A comparison of the homozygous and heterozygous pie mice revealed that the ectopia distributions were significantly different from each other (X z = 11.04, df = 1, P < 0.001).

3.3. Cloning of the transgene insertion site The cloning of a junction fragment containing both transgene and wild-type genomic sequences is described in Section 2. Characterization of the genomic region on this junctional clone was complicated by the presence of abundant mouse repetitive sequence. DNA sequencing was utilized to identify a 300 bp region that, when used as a probe, hybridizes to a single band on a Southern blot that is polymorphic between the transgene and wild-type DNA. Since this probe, called Dl5Beil in accordance with the rules for the nomenclature of anonymous DNA clones [ 18], can unambiguously distinguish wild-type, heterozygous,

201

and homozygous transgenic animais, it has been useful for genotyping mice (Fig. 1). The Dl5Beil probe was used to identify hybridizing clones in a genomic library prepared in EMBL3. Positive clones were identified, and one C~4) was proven to contain sequences corresponding to Dl5Beil by sequence analysis. Subcloned fragments and PCR-based probes derived from a region of this clone that lies 'downstream' from the transgene junction site fail to hybridize to genomic DNA prepared from homozygous ple mice. demonstrating that the ple transgene insertion is associated with a deletion of wild-type genomic DNA (Fig. 1). Probes from the deleted region were used to extend the region of cloned sequences from wild-type DNA; the sequences contained on one contiguous bacteriophage clone (~3) are apparently entirely deleted from pie mice. The extent of this deletion is as yet undefined, since we bave hot obtained mouse sequence corresponding to the second junction of the transgene and wild-type sequence. It appears at least greater

Fig. 3. (A-D) Digitized composite (in Adobe Photoshop) showing (A) Nissl staining of a large ectopia (arrow) in adult pie homozygous mouse: (B) adjacent section stained with a monoclonal antibody directed against neurofilament protein (68 kDa). Note the increased density of the oeurofilament fibers (black arrowheads) underlying the ectopia. (C) Nissl stain of a large ectopia (arrow) in a ferai pie homozygous mouse, Note the underlying distortion of cortical layers (white arrows); (D) adjacent section to panel C stained with a monoclonal antibody (Rat-401) directed against radial glial fibers. Note the breach in the glial-pial border (white arrows). Processing artifacts were digitally excised. Bar 250 jxm for A / B and 65 gin for C / D

202

D.R. Beier et al. / Brain Research 727 (1996) 196 204

A

C

I ~

' ~/'--.ddm~

ID

Fig. 4. ( A - D ) Digitized composite (in Adobe Photoshop) showing Nissl stained coronal sections containing 4 layered microgyria (arrowheads), ectopia (arrow) and related gross distortion of the right hemisphere (right is left side) of a newborn pie homozygous/ple mouse. Note the presence in panel D of the small ectopic collection of neurons (arrow) in layer i of the opposite hemisphere. Panels are arranged rostral to caudal with approximately 0.1 mm between sections. Processing artifacts were digitally excised. Bar = 250 p,m.

than 16 kb, which is the length of the X3 bacteriophage clone,

4. Discussion

In an inbred genetic background, mice homozygous for a transgene insertion at ple were found to be significantly smaller than their heterozygous or wild-type siblings at birth, and rarely survived for more than 48 h. When C57BL/6J pie mice were crossed with mice from the strain C3Heb/FeJLe, homozygous progeny of the F1 mice were viable, but were still not obtained in the expected numbers, demonstrating some deleterious effect of this presumptive mutation even in the hybrid genetic background. Examination of these homozygous mice revealed variable expression of abnormalities in brain development. These usually appeared as focal cerebrocortical ectopias, but also included other related abnormalities, such as microgyria and gross distortion of cortical layering. The

ectopias are similar to those that prenatally develop in N Z B / B I N J and B X S B / M p J mice [25,26], although they were generally larger in the ple mutants. Similar ectopias and microgyria can also be induced at birth in rats [9,15,23]. Since the transgene is an active MMTV/murine c-my« chimeric gene, containing an intact c-myc promoter, i t i s not yet possible to prove that the phenotype in homozygous transgenic mice is not due to transgene dosage. However, the early onset of mortality and the absence of similar phenotype in numerous transgenic lines carrying MMTV/c-myc argues against a transgene dosage effect. In addition, the identification of a significant deletion of genomic sequences at the transgene insertion site supports the likelihood tbat this insertion is associated with a disruption of an endogenous locus. The phenotype of ple is hOt similar to other mutations mapped near the Ca locus and therefore probably represents a new mutation. While the exact nature of the mechanism of neuronal migration is not entirely known, there is considerable support for the hypothesis that many neurons achieve their

D.R. Beier et a l . / B r a i n Research 727 (1996) 196-204

final position in the cortical plate by migrating out of the ventricular zone along a fascicle of radial fibers [13,21]. In some cases of disordered neuronal migration, such as the weaver and r«eler mutations, the defect appears to be intrinsic to the neuron [12,22]. Analysis of the expression of reelin (the reeler gene product) by in situ hybridization [8] or antibody analysis [19] support a neuronal-specific effect. How this molecule, which bas motifs suggesting that i t i s either secreted or bound to the external cell surface, specifies neuronal localization remains to be determine& In the ple mutation described here, however, and in the NZB and BXSB mouse strains, it is perhaps more likely that the defect is in the external glial limiting membrane where the endfeet of the radial glia are attached. This supefficial cortical boundary must be intact for normal neuronal migration to take place. It has been suggested that damage to this border could result in the production of focal cortical layer I ectopias seen in human autopsy specimens [5-7] and in NZB and BXSB mice [27]. There are a variety of potential mechanisms by which the hypothesized defect in the external glial limiting membrane may occur. One possibility would be as a consequence of a local injury, such as bleeding or vascular occlusion, that results in ischemic damage to the cortical boundary. This model is consistent with the usually focal nature of the abnormality in pie mice, its variable expression, and the evidence that similar lestons can be induced in the developing rat brain by cryogenic or puncture injury [9,15,23]. Depending on the size of the ischemic injury and timing of the insult, the resultant damage to the developing cerebral cortex could be either gross distortion of layering, microgyria or molecular layer ectopias. An alternative model would be that there is a primary defect in the development of an intact cortical boundary. Another possibility is suggested by the evidence that reeler, whose phenotype includes the abnormal localization of neurons in the marginal zone of the fore-brain cortex [22], is a defect in a neuronal-specific gene. And, while the focal nature of the abnormalities found in pie mice might argue against the hypothesis that its defect is neuron-specific, it has recently been shown in a transgenic system that selective gene expression in specific cortical radial units may occur, suggesting there can be molecular heterogeneity of cortical progenitors [29]. The identification of the gene disrupted by the pie transgene insertion, and characterization of its pattern of expression and function, should facilitate a more specific examination of these questions. It should be noted that, while the ectopias in the pie and NZB mite appear similar, it is unlikely that they are due to an identical molecular abnormality. This is suggested by thc fact that. in an inbred genetic background, pie mice are mnted and usually do not survive the perinatal period: that is, the cortical ectopias seen in these mice are likely part of a more global defect. The presence of ectopias is not itself incompatible with survival, as the incidence of ectopias in NZB mice is, at 40%, higher than

21)3

that found in pie mice, and the incidence in the NXSM recombinant inbred subline D is higher yet, at 80% [28]. In addition, NZB mice rarely have microgyria. Finally, an analysis of NZB mice showed no evidence of genetic linkage of the region of chromosome 15 identified by the transgene with the ectopia phenotype (unpublished results). In summary, we have found that a transgenic insertional mutation that results in perinatal lethality in inbred mite is associated with a focal abnormality of cortical neuronal migration. Transgenic insertions have proved extremely useful as molecular 'tags' that can be utilized for the cloning of disrupted loci; examples that bave been successfully characterized include insertional mutations in limbdeformity (ld) [31 ], mic;'o-ol?thalmia (mi) [ 14], and reeler [8]. We have cloned a junction fragment containing the pic transgene and the adjacent mouse genomic sequences, and used this fragment to clone the lotus of the transgene insertion from wild-type mouse DNA (data not shown). Il should be feasible to apply recently developed strategies for the identification of transcribed sequences in cloned genomic DNA, such as cDNA selection [17,20] and exontrapping [3]. for the purpose of identifying the gene disrupted by the insertion at ple that is apparently required for normal cortical development.

Acknowledgements We are grateful for the encouragement and support of Dr. Philip Leder, in whose laboratory this work was initiated. We thank Dr. Robert D. Cardiff and the University of California at Davis Transgenic Histophatology Laboratory for their help. The Rat-401 antibody was obtained from the Developmental Studies Hybridoma Bank maintained by the Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine, Baltimore. MD 21205 and the Department of Biologica[ Sciences, University of Iowa, Iowa City, lA 52242, under contract NO1-HD-2-3144 from the NICHD. This work was supported by grant RO1-HD29028 from the National Institute of Child Health and Development fo D.R.B. and NICHD 20806 to G.F.S.

References [I] Barth, P.G., Disorders of neuronal miuation, Ca;l. ,/. ,,V«uro/. SeL, 114(I987) 1-16. [2] Beier, D.R., Morton, C.C,, Lcd«r, A.. Wallacc. R. and Leder, P.. Perinatal lethality (pie): a mutation caused bv inlegralion of a transgene into distal mouse chromoson3e 15. G«mn,i«v, 4 (1989) 498 504. [3] Buckler, A.J., Chamg, D.D., Grav,, S.E., Brook, .I.D.. Haber. D.A., Sharp, P.A, and Housman, D.E., Exon amplification: a strategy 1o isolate mammalian genes based on RNA splicing, l'roc. Natl. A«ad. Soi. U&4, 88 ( 1991 ) 4005-4009. [4] Caviness, V.S.. Neocortical histogenesis and recler mite: a devclopmental study based upon ( ~H)thymidine auloradiography, l)e;'. Brait; Res,, 4 (1982) 293-302.

204

D.R. Beier et al. / Brain Research 727 (1996) 196-204

[5] Caviness, V.S., Evrard, P. and Lyon, G., Radial neuronal assemblies, ectopia and necrosis of developing cortex: a case analysis, Acre Neuropathol., 41 (1978) 67-72. [6] Caviness, V.S., Misson, J.-P. and Gadisseux, J.-F., Abnormal neuronal patterns and disorders of neortical development. In A.M. Galaburda (Eds.), From Reading to Neurons, MIT Press/Bradford Books, Cambridge, MA, 1978, pp. 405-442. [7] Dambska, M., Wisniewski, K.E. and Sher J.H., Marginal glioneuronal heterotopias in nine cases with and without cortical abnormalities, J. Child N«urol., 1 (1986) 149-157. [8] D'Arcangelo, G., Miao, G.G., Chen, S.-C., Soares, H.D., Morgan, J.l. and Curran, T., A protein related to extracellular matrix proteins deleted in the mouse mutant reeler. Nature, 374 (1995) 719-723. [9] Dvorak, K. and Feit, J., Migration of neuroblasts through partial necrosis of the cerebral cortex in newborn rats - contribution to the problems of morphological development and developmental period of cerebral microgyria, Acre Neuropathol., 38 (1977) 203-212. [10] Farrell, M.A., DeRosa, M.J., Curran, J.G., Secor, D.L., Cornford, M.E., Comair, Y.G., Peacock, W.J., Shields, W.D. and Vinters, H.V., Neuropathologic findings in cortical resections (including hemispherectomies) performed for the treatment of intactable childhood epilepsy, Acta N«uropathol., 83 (1992) 246-259. [11] Galaburda, A.M., Sherman, G.F., Rosen, G.D., Aboitiz, F. and Geschwind, N., Developmental dyslexia: four consecutive patients with cortical anomalies, Ann. Neurol., 18 (1985) 222-233. [12] Gao, W,Q. and Hatten, M.E.. Neuronal differentiation rescued by implantation of Weaver granule cell precursors into wild-type cerebellar cortex, Science, 260 (19931 367-369. [13] Gressens, P. and Evrard, P., The glial fascicle: an ontogenic and phylogenetic unit guiding, supplying and distributing marnmalian cortical neurons, Del,. Brain Res., 76 (1993) 272 277. [14] Hodgkinson, C.A., Moore, K.J., Nakayama, A., Steingrimsson, E., Copeland, N.G., Jenkins. N.A. and Arnheiter. H., Mutations at the mouse micropthalmia locus are associated with defects in a gene encoding a novel basic-helix-loop-helix-zipper protein, Cell, 74 (1993) 395-404. [15] Humphreys, P., Rosen, G.D,, Press, D.M., Sherman, G.F. and Galaburda, A.M., Freezing lesions of the newborn rat brain: a model for cerebrocortical microgyria, J. Neuropathol. Exp. Neurol., 50 (1991) 145-160. [16] Leder, A., Pattengale, P.K., Kuo, A., Stewart, T.A. and Leder, P., Consequences of widespread deregulation of the c-myc gene in transgenic mice: multiple neoplasms and normal development, Cell, 45 (1986) 485-495. [17] Lovett, M., Kere, J. and Hinton, L.M., Direct selection: a method for the isolation of cDNAs encoded by large genomic regions, Proc. Natl. Acad. Sci, USA, 88 (199t) 9628-9632. [18] Lyon, M.F. and Searle, A.G. (Ed.), Genetic Variants and Strains of

the Lahoratorv Mouse (2nd ed.), Oxford University Press. Oxford. 1989. [19] Ogawa. M., Miyata, T., Nakajima, K., Yagyu, K., Seike, M., lkenaka, K., Yamamoto, H. and Mikoshiba K. The reeler gene-associated antigen on Cajal-Retzius neurons is a crucial molecule for laminar organization of cortical neurons, Neuron. 14 (1995) 899912. [2(I] Parimoo, S., Patajal, S.R., Shukla, H., Chaplin, D.D. and Weissman, S.M., cDNA selection: efficient PCR approach for the selection of cDNAs encoded in large chromosomal DNA fragments, Pro«. Natl. A«ad. SeL USA, 88 (1991) 9623-9627. [21] Rakic, P.. Radial versus tangential migration of neuronal clones in the developing cerebral cortex, Proc. Natl. A«ad. Soi. USA, 92 (,19951 11323 11327. [22] Rakic, P. and Caviness, V.S.. Cortical development: view from neurological mutants two decades later, Cell, 14 (19951 I 101 1104, [23] Rosen. G.D.. Shennan, G.F., Richman, J.M.. Stone, LV. and Galaburda. A.M., Induction of molecular layer ectopias by puncture wounds in newborn rats and mice, DeL,. Bruin Res., 67 (1992) 285-291. [24] Sambrook, J., Fritsch, E.F. and Maniatis. T., Mole«ular Cloning (2 ed,), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989. [25] Sherman, G.F., Galaburda, A.M., Behan, P.O. and Rosen, G.D,, Neuroanatomical anomalies in autoimmune mice, A«tu N«uropathol., 74 (1987) 239-242. [26] Sherman. G.F., Galaburda, A.M. and Geschwind, N., Cortical anomalies in brains of New Zealand mice: a neuropathologic model of dyslexia? Pro«. Natl. Acud. Sci. USA, 82 (1985) 8072-8074. [27] Sherman, G.F., Rosen, G.D., Stone, L.V., Press, D.M. and Galaburda, A.M., The organization of radial glial fibers in spontaneous neocortical ectopias of newborn New Zealand Black mice, Der. Brain Res., 67 (1992) 279-283. [28] Sherman, G.F., Stone, L.V., Denenberg, V.H. and Beier, D.R., A genetic analysis of neocrotical ectopias in New Zealand Black autoimmune mice, Neuroreport, 5 (1994) 721-724. [29] Soriano, E., Dumesnil, N., Auladell, C., Cohen-Tannoudji M., Sotelo, C., Molecular heterogeneity of progenitors and radial migration in the developing cerebral cortex revealed by transgene expression, Pro«. Nutl. A«ad, S«i. USA, 92 (1995) 11676-80. [30] Stewart. T.A., Pattengale, P.K. and Leder, P., Spontaneous mammary adenocarcinomas in transgenic mice that carry and express MTV/myc Iusion genes, Cell, 38 (1984) 627-637. [31] Woychik, R.P., Stewart, T.A., Davis, L.G.. D'eustachio, P. and Leder, P., An inherited limb deformity created by insertional mutagenesis in a transgenic mouse, Nature, 318 (1985) 36-40. [32] Zilles. K., Th« Cortex o1" the Rat, Springer-Verlag, Berlin. 1985.